Push dataset with binning and relative time filtering

[Anonymous]

Hi Folks, I am trying to create a graph of Transaction Count per 15min time interval for the last 6 hours of transactions. The intention is that this graph will be on a dashboard to make use of the push dataset auto updating feature

I am trying to run this on a Push Dataset table that includes the following fields

• Transaction ID (string)
• Created Date (datetime)
• Amount (decimal)

Graphing [Created Date] vs COUNT([Transaction ID] gets me some of the way there but have two issues

1. How do I bucket my dates into 15min intervals. The binning feature does not seem to work for push datasets.
2. How do a do a relative time filter. The filters built in can only filter by whole days, not hours.

This seems like it would be a standard graph for push (and streaming) datasets so I am hoping there is just some feature I am not seeing

[ibarrau · Accepted solution]

Hi. I'm not sure about the first one but the second one is now release as preview in last weeks:

https://powerbi.microsoft.com/en-us/blog/power-bi-desktop-april-2020-feature-summary/#_Relative_time...

Please update your power bi desktop to work with that.

Before that update what you could have used was filtering your result by "CALCULATE ( [Measure], Table[DateColumna] > NOW() - 0,0104)"

That will take last 15 minutes for example.

Regards

[Anonymous]

It was fortutious that they released the relative time filter in the most recent update. Thanks for pointing it out that makes it a lot easier. The alternative way to do it seems considerably less flexiable

As for the first point I believe I can aggregate it into 15min slots at the source and add a new 'Date Created 15min timeslice' field. It is not very flexiable but can be workable